Continuous production process optimization using machine learning

Abstract

One embodiment of the invention provides a computer-implemented method for optimization of a continuous production process. The method comprises receiving input data comprising a plurality of datasets each including one or more variables relating to a production equipment involved in the continuous production process. The method further comprises generating different prediction models based on the input data. Each of the prediction models is configured to output a target prediction relating to the production equipment. The method further comprises generating an objective optimization model based on each target prediction output from each of the prediction models. The objective optimization model comprises a deep neural network. The method further comprises generating a loss function corresponding to the objective optimization model, and optimizing weights for parameters of the prediction models using backpropagation of the deep neural network and the loss function, resulting in optimized weights for the parameters of the prediction models.

What is claimed is: 
     
         1 . A computer-implemented method for optimization of a continuous production process, comprising:
 receiving input data comprising a plurality of datasets each including one or more variables relating to a production equipment involved in the continuous production process;   generating different prediction models based on the input data, wherein each of the different prediction models is configured to output a target prediction relating to the production equipment;   generating an objective optimization model based on each target prediction output from each of the different prediction models, wherein the objective optimization model comprises a deep neural network;   generating a loss function corresponding to the objective optimization model; and   optimizing a plurality of weights for a plurality of parameters of the different prediction models using backpropagation of the deep neural network and the loss function, resulting in a plurality of optimized weights for the parameters of the different prediction models.   
     
     
         2 . The computer-implemented method of  claim 1 , further comprising:
 splitting the datasets between a first group of datasets for training and a second group of datasets for testing;   defining a problem type for each of the different prediction models;   selecting a machine-learning algorithm for training each of the different prediction models; and   selecting one or more statistical measures for evaluating performance of each the different prediction models.   
     
     
         3 . The computer-implemented method of  claim 1 , wherein the datasets include different variables with different time ranges that correspond to the different prediction models. 
     
     
         4 . The computer-implemented method of  claim 1 , further comprising:
 providing, as output, the optimized weights for the parameters of the different prediction models.   
     
     
         5 . The computer-implemented method of  claim 1 , further comprising:
 providing, as output, the different prediction models with fixed parameters based on the optimized weights.   
     
     
         6 . The computer-implemented method of  claim 1 , further comprising:
 generating an objective function corresponding to the objective optimization model, wherein the objective function represents one or more production optimization goals; and   generating constraints corresponding to the objective optimization model.   
     
     
         7 . The computer-implemented method of  claim 6 , further comprising:
 training the objective optimization model to minimize a difference quantified by the loss function, wherein the difference is between an expected objective value and a predicted objective value, the expected objective value is based on the objective function, and the predicted objective value is output from the objective optimization model.   
     
     
         8 . The computer-implemented method of  claim 1 , wherein an input layer of the deep neural network propagates initial weight matrices representing configurations of the different prediction models. 
     
     
         9 . A system for optimization of a continuous production process, comprising:
 at least one processor; and   a processor-readable memory device storing instructions that when executed by the at least one processor causes the at least one processor to perform operations including:
 receiving input data comprising a plurality of datasets each including one or more variables relating to a production equipment involved in the continuous production process; 
 generating different prediction models based on the input data, wherein each of the different prediction models is configured to output a target prediction relating to the production equipment; 
 generating an objective optimization model based on each target prediction output from each of the different prediction models, wherein the objective optimization model comprises a deep neural network; 
 generating a loss function corresponding to the objective optimization model; and 
 optimizing a plurality of weights for a plurality of parameters of the different prediction models using backpropagation of the deep neural network and the loss function, resulting in a plurality of optimized weights for the parameters of the different prediction models. 
   
     
     
         10 . The system of  claim 9 , wherein the operations further include:
 splitting the datasets between a first group of datasets for training and a second group of datasets for testing;   defining a problem type for each of the different prediction models;   selecting a machine-learning algorithm for training each of the different prediction models; and   selecting one or more statistical measures for evaluating performance of each the different prediction models.   
     
     
         11 . The system of  claim 9 , wherein the datasets include different variables with different time ranges that correspond to the different prediction models. 
     
     
         12 . The system of  claim 9 , wherein the operations further include:
 providing, as output, the optimized weights for the parameters of the different prediction models.   
     
     
         13 . The system of  claim 9 , wherein the operations further include:
 providing, as output, the different prediction models with fixed parameters based on the optimized weights.   
     
     
         14 . The system of  claim 9 , wherein the operations further include:
 generating an objective function corresponding to the objective optimization model, wherein the objective function represents one or more production optimization goals; and   generating constraints corresponding to the objective optimization model.   
     
     
         15 . The system of  claim 14 , wherein the operations further include:
 training the objective optimization model to minimize a difference quantified by the loss function, wherein the difference is between an expected objective value and a predicted objective value, the expected objective value is based on the objective function, and the predicted objective value is output from the objective optimization model.   
     
     
         16 . A computer program product for optimization of a continuous production process, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:
 receive input data comprising a plurality of datasets each including one or more variables relating to a production equipment involved in the continuous production process;   generate different prediction models based on the input data, wherein each of the different prediction models is configured to output a target prediction relating to the production equipment;   generate an objective optimization model based on each target prediction output from each of the different prediction models, wherein the objective optimization model comprises a deep neural network;   generate a loss function corresponding to the objective optimization model; and   optimize a plurality of weights for a plurality of parameters of the different prediction models using backpropagation of the deep neural network and the loss function, resulting in a plurality of optimized weights for the parameters of the different prediction models.   
     
     
         17 . The computer program product of  claim 16 , wherein the program instructions executable by the processor further cause the processor to:
 split the datasets between a first group of datasets for training and a second group of datasets for testing;   define a problem type for each of the different prediction models;   select a machine-learning algorithm for training each of the different prediction models; and   select one or more statistical measures for evaluating performance of each the different prediction models.   
     
     
         18 . The computer program product of  claim 16 , wherein the datasets include different variables with different time ranges that correspond to the different prediction models. 
     
     
         19 . The computer program product of  claim 16 , wherein the program instructions executable by the processor further cause the processor to:
 provide, as output, the optimized weights for the parameters of the different prediction models.   
     
     
         20 . The computer program product of  claim 16 , wherein the program instructions executable by the processor further cause the processor to:
 provide, as output, the different prediction models with fixed parameters based on the optimized weights.